Detect if text was written by ChatGPT, Gemini, Claude or other AI models. Analyzes perplexity, burstiness, and linguistic patterns. 100% private — runs in your browser.
0 chars · 0 words
📄 Text Analysis View
Purple = likely AI · Green = likely human · Click sentences to inspect
🔮
Enter text and click Analyze to get your AI detection score.
How it works: We analyze vocabulary repetition, sentence uniformity (burstiness), rare word ratio, punctuation patterns, and structural clichés typical of AI-generated text.
Human 0%AI 100%
Sentence-level breakdown
📊 Signal Breakdown
Frequently Asked Questions
This tool uses heuristic analysis of burstiness, perplexity proxies, vocabulary diversity, and AI-typical phrasing patterns. It's most accurate on texts over 100 words. No detector is 100% accurate — results should be used as a guide, not a definitive verdict.
Yes. All modern LLMs (ChatGPT, GPT-4, Gemini, Claude, Copilot) share similar statistical patterns — uniform sentence length variance, low burstiness, high predictability, and typical AI filler phrases. Our detector targets these patterns across all AI systems.
No. All analysis runs entirely in your browser using JavaScript. Your text is never sent to any server and is never stored. Complete privacy guaranteed.
Burstiness measures how varied sentence lengths are. Humans naturally write with high variation — very short punchy sentences mixed with long complex ones. AI models tend to produce uniformly medium-length sentences, which shows up as low burstiness — a key AI signal.